Abstract Background and aims Carotid near-occlusion (CNO) is a severe stenosis causing distal artery collapse. Symptomatic CNO is recommended for conservative treatment, but accurate CNO diagnostics requires expert assessment. This study aimed to develop diagnostic criteria for computed tomography angiography (CTA) to optimize CNO diagnostics for those not well-acquainted with more advanced CNO diagnostic techniques. Methods Cross-sectional study of cases with ≥50% stenosis. The CNO reference test was performed by expert assessment of CTA, whereas CTA measurements were collected by blinded, trained, and inexperienced assessors. Several variants of diagnostic analyses were performed, including a pragmatic rule-in/rule-out approach with a requirement for 98% sensitivity and 99% specificity. Here, the fewest possible uncertain cases (neither ruled in or out) were sought. Results We included 459 cervical sides with ≥50% stenosis from 369 patients. With a traditional analysis approach, no measurement (single or combined) was sufficiently accurate for clinical decision-making, with the best performance having a sensitivity and specificity 90%, with a positive predictive value of 83%. With a pragmatic approach, a scoring system of four measurements resulted in 26% (95% CI 22-30%)uncertain cases. Conclusions While measurements collected by non-experts cannot be used to diagnose all cases as either CNO or not CNO, a pragmatic CNO score seems to be able to diagnose or exclude CNO in a large share of cases. This method still left a quarter of cases needing further assessment with more advanced methods. Conflict of interest Anna Öjendal: nothing to disclose. Alexander Henze: nothing to disclose. Allan Fox: nothing to disclose. Elias Johansson: nothing to disclose.
Öjendal et al. (Fri,) studied this question.